Recovering large sets of digitally-stored data is an increasingly important task for organizations. To protect against data loss, an organization may use a backup system to back up important data. Traditional backup systems may generate recovery log files during each incremental backup that indicate which portions of a data set have changed during the incremental backup. Such backup systems may then later use these incremental recovery log files to recreate the data set as it was at a previous time. For example, a conventional backup technology may synthesize a previous version of a data volume based on information indicating which blocks within the data volume have changed since the previous version was created.
Unfortunately, these techniques may be ineffective and/or inefficient when applied to certain types of files or data. For example, creating full backups and/or synthesizing previous versions of data sets (e.g., by updating and/or reverting data blocks at a file system level) may only be successful for data volumes (e.g., databases). Such techniques may not work for applications, as applications may store data in multiple files across multiple levels of storage. The instant disclosure, therefore, identifies and addresses a need for improved systems and methods for generating full backups of applications.